Occupation-Level Variance

Occupation codes changed between 2010 and 2018 Census vintages. This analysis harmonizes both into a single frame and examines within-occupation earnings dispersion by gender across ~500 occupations and 11 ACS years (2013–2024, pooled 9.7M observations).

Sample window note: This page pools ACS 2013–2024 to maximize occupation cell sizes. The headline trend series covers 2015–2023. Results here are a separate pooled extension, not directly comparable to the year-by-year adjusted gap estimates.

How to Read This

Variance ratio > 1: male earnings are more dispersed than female in that occupation.
Variance ratio < 1: female earnings are more dispersed than male.
Top-decile gap < 0: men account for a larger share of the top earnings decile.
Top-decile gap > 0: women account for a larger share of the top earnings decile.
All variance ratios are male / female (M/F), matching the Variance page and standard literature convention.
All results are descriptive summaries of the earnings distribution, not causal estimates. Small cell sizes (marked N<1K) increase noise in individual occupation estimates.

Harmonization types: 2018 native 2018 code   mixed 1:1 crosswalk   split 1-to-many split bucket   2010 legacy 2010 only

Variance Leaderboards

Female-Higher-Variance Occupations (Annual)

Variance ratio (M/F) — lowest 10 (female more dispersed)

Female earnings are more dispersed than male earnings in these occupations. Source: results/variance/acs_occupation_variability_leaders.csv

OccupationRatio (M/F)ResidualSOCNType
Agents/biz mgrs of artists0.130.06Business/Fin.2,603mixed
Misc. vehicle/mobile equip mechanics0.210.23Install./Repair5,011mixed
Elevator & escalator installers0.250.25Construction2,085mixed
Parts salespersons0.260.19Sales7,017mixed
Solar photovoltaic installers0.270.33Construction6452018 N<1K
Metal workers, all other (2010)0.270.25Production13,3922010
Food/tobacco machine operators0.390.42Production719mixed N<1K
Electrical power-line installers0.400.43Install./Repair9,850mixed
Furnace/kiln/oven operators0.410.32Production842mixed N<1K
Forming machine operators0.440.51Production1,1822018

Male-Higher-Variance Occupations (Annual)

Variance ratio (M/F) — top 10 (male more dispersed)

Male earnings are more dispersed than female earnings in these occupations. Source: results/variance/acs_occupation_variability_leaders.csv

OccupationRatio (M/F)ResidualSOCNType
Child/family/school social workers2.082.33Community/Social3,2162018
Rehabilitation counselors2.041.89Community/Social9522018 N<1K
Production helpers (2010)2.041.92Production1,271split
Avionics technicians1.921.54Install./Repair1,518mixed
Mental health/substance abuse social workers1.821.52Community/Social1,1162018
Title examiners/abstractors1.821.41Legal3,4022018
Paralegals and legal assistants1.821.69Legal30,142mixed
Photographers1.792.70Arts/Design5,416mixed
Passenger attendants1.691.61Transportation7802018 N<1K
Fundraisers1.671.59Business/Fin.7,095mixed

Female-Higher-Variance Occupations (Hourly)

Variance ratio (M/F) — lowest 10 (female more dispersed)

Female hourly wages are more dispersed than male in these occupations.

OccupationRatio (M/F)ResidualSOCNType
Agents/biz mgrs of artists0.250.11Business/Fin.2,603mixed
Cement masons/concrete finishers0.300.27Construction3,882mixed
Elevator & escalator installers0.350.35Construction2,085mixed
Millwrights0.370.42Install./Repair3,715mixed
Metal workers, all other (2010)0.370.29Production13,3922010
Food/tobacco machine operators0.380.44Production719mixed N<1K
Tool and die makers0.420.43Production3,052mixed
Helpers, install./repair0.430.43Install./Repair6282018 N<1K
Electrical/electronics engineers0.430.47Architecture/Eng.17,223mixed
Other personal appearance workers0.470.45Personal Care5912018 N<1K

Male-Higher-Variance Occupations (Hourly)

Variance ratio (M/F) — top 10 (male more dispersed)

Male hourly wages are more dispersed than female in these occupations.

OccupationRatio (M/F)ResidualSOCNType
Photographers2.864.17Arts/Design5,419mixed
Production helpers (2010)2.632.27Production1,271split
Helpers, install./repair (2010)2.562.17Install./Repair577split N<1K
Medical transcriptionists2.221.92Healthcare Supp.2,127mixed
Control/valve installers2.172.27Install./Repair1,614mixed
Explosives workers/blasters2.172.78Construction6352018 N<1K
Library assistants, clerical2.131.92Office/Admin.5,638mixed
Marine engineers/naval architects2.082.13Architecture/Eng.1,084mixed
Fishing and hunting workers1.962.70Farming/Fishing6602018 N<1K
Credit analysts1.961.96Business/Fin.2,485mixed

Female earnings are more dispersed in occupations concentrated in production, construction, and installation/repair. Male earnings are more dispersed in occupations concentrated in community/social service, legal, and arts fields. The pattern is consistent with the minority gender in a given occupation tending to show wider earnings spread, though other factors may also contribute.

Top-Decile Concentration

Largest Male Top-Decile Advantage (Annual)

Occupations where men account for the largest share of top-decile earnings

Gap = female top-10% share − male top-10% share (pp). More negative = larger male advantage.

OccupationF Top-10%M Top-10%Gap (pp)N
Probation officers / correctional0.8%76.1%−75.37,560
Parts salespersons4.3%77.7%−73.47,017
Photographers3.6%37.3%−33.75,416
Dental hygienists9.1%32.3%−23.115,322
Diagnostic medical sonographers6.3%26.2%−20.04,448
Pressers, textile/garment3.2%22.3%−19.12,198
Credit authorizers/checkers5.0%23.2%−18.22,828
Title examiners/abstractors5.2%23.1%−17.93,402
Production helpers (2010)2.3%19.4%−17.11,271

Largest Female Top-Decile Advantage (Annual)

Occupations where women account for the largest share of top-decile earnings

Gap = female top-10% share − male top-10% share (pp). More positive = larger female advantage.

OccupationF Top-10%M Top-10%Gap (pp)N
Agents/biz mgrs of artists70.1%5.7%+64.42,603

Only one occupation in the top-25 absolute-gap leaderboard shows a female top-decile advantage. This asymmetry is itself a descriptive finding: across most occupations with large absolute top-decile gaps, men account for a larger share of top earners than women.

Largest Male Top-Decile Advantage (Hourly)

Occupations where men account for the largest share of top-decile hourly wages

OccupationF Top-10%M Top-10%Gap (pp)N
Probation officers / correctional1.0%76.1%−75.17,560
Parts salespersons5.1%77.2%−72.17,017
Photographers4.4%33.3%−28.85,419
Title examiners/abstractors5.5%22.4%−16.93,402
Dental hygienists9.5%25.3%−15.815,322
Diagnostic medical sonographers7.1%22.5%−15.44,448
Pressers, textile/garment4.7%19.9%−15.22,198
Credit authorizers/checkers5.9%20.7%−14.72,828
Production helpers (2010)3.4%18.0%−14.61,271

Largest Female Top-Decile Advantage (Hourly)

Occupations where women account for the largest share of top-decile hourly wages

OccupationF Top-10%M Top-10%Gap (pp)N
Agents/biz mgrs of artists72.0%5.5%+66.52,603

As with annual earnings, only one occupation in the top-25 by absolute gap shows a female advantage in top-decile concentration. In the vast majority of occupations with large absolute gaps, men hold a larger share of the top decile.

SOC Group Concentration

Top-10 Leaderboard Seats by SOC Major Group

Count of occupations in each top-10 leaderboard, by industry group

Each bar = number of occupations from that SOC group appearing in the given top-10 leaderboard. Annual earnings outcome.

SOC GroupFemale HigherMale Higher
Production41
Construction & Extraction20
Installation / Repair21
Community & Social Service03
Legal02
Business / Financial11
Sales10
Arts / Design / Media01
Transportation01

Female-higher-variance occupations are concentrated in production, construction, and installation/repair. Male-higher-variance occupations are concentrated in community/social service and legal fields. One possible interpretation is that the gender in the minority within an occupation tends to show wider earnings spread, but other factors — including occupational structure and measurement — may also play a role.

Pre/Post-2020 Variance Regime

Hourly Variance Metrics: Pre-2020 vs. Post-2020

Weighted means across ACS years in each regime

Pre-2020: ACS 2013–2019 (N=6.1M). Post-2020: ACS 2021–2024 (N=3.6M). No 2020 ACS available.

Metric (Hourly)Pre-2020Post-2020Direction
Raw variance ratio (M/F)1.1221.085toward parity
Residual variance ratio (M/F)1.0531.054stable
Female top-10% share7.47%7.82%increased
Male top-10% share12.65%12.29%decreased
Female top-5% share3.47%3.75%increased
Male top-5% share6.67%6.19%decreased

Post-2020, hourly dispersion moved modestly toward parity: the within-occupation M/F variance ratio decreased from 1.12 to 1.08, and women gained slightly in top-earner shares. The residual ratio was essentially unchanged. These shifts are consistent with a mild post-pandemic compression of hourly extremes rather than a structural break.

Methods & Caveats

Occupation harmonization: Census occupation codes changed in 2018. This analysis maps all years onto a unified frame using native 2018 codes, 1:1 crosswalks, split buckets, and legacy-2010-only buckets. The full mapping is in results/diagnostics/variance_occupation_harmonization_map.csv.

Small cell sizes: Some leaderboard occupations have fewer than 1,000 pooled observations (marked N<1K). Variance ratios in these cells are noisier — treat individual rankings as indicative.

Fertility-risk bridge: A descriptive bridge between fertility-risk quartiles and variance metrics is available in results/variance/acs_fertility_risk_variance_bridge.csv. It has weaker male/female ratio coverage and is not charted here.

Interpretation: All results are descriptive summaries of observed earnings distributions. Higher or lower variance within an occupation could reflect differences in tenure, specialization, hours composition, hiring patterns, or measurement. These tables identify where dispersion differs, not why.